GEN-base and reuse

From Logius SBR EN Wiki
Jump to: navigation, search

Reuse of terms

Reuse of data is fundamental to SBR when creating Dutch Taxonomy domain extensions. Element definition is, as far as possible, based on existing and widely supported (government) standards. The aim of the design and organisation of the Dutch Taxonomy is to facilitate maximum re-use. This page describes how the re-use of data in the Dutch Taxonomy is shown. It describes the way in which a Dutch Taxonomy domain extension is made from a technical point of view here.

The Dutch Taxonomy contains two specific taxonomies which express the widely shared terms: the NL-CD and NL-GEN. The NL-CD and NL-GEN are, jointly, also referred to as the 'GEN-base'. GEN stands for Generic Elements Netherlands. The NL-CD (Dutch Common Dictionary) contains all shared non-financial terms; for example, name, address and identification numbers. The NL-GEN contains the generic financial terms that appear in almost all financial reports, such as the balance sheet, profit & loss report and information that is commonly requested.


The Dutch Taxonomy partners have decided to modify the design of the NL-GEN as from Dutch Taxonomy version 6.0 (Dutch Taxonomy 2012). In the NL-GEN, the so-called “downward principle” no longer exists, whereby elements that are used by at least two Dutch Taxonomy partners automatically appear in the Dutch Taxonomy for the subsequent year in the NL-GEN taxonomy. Through the “downward principle”, data is only reused by inclusion of these terms in the NL-GEN. This model is found to be too restrictive by the parties creating the taxonomy and it is not thought to be future-proof.It is anticipated that the main thrust of the forthcoming period will be the connection of more and new domains within the scope of the SBR expansion. With a large group, a great deal of consultation and coordination is essential until this results in a final NL-GEN, which impacts the schedule. For that reason, the “downward principle” has been replaced by a hybrid model that aims at increased reuse of data between the various parties creating the taxonomy. Under this hybrid model there are three options for reusing data:

1. The use of the generic terms from the NL-GEN;
2. Using one another's terms directly in the domain taxonomies;
3. Copying a concept of another party that is creating a taxonomy to use in own domain taxonomy;

Unlike the “downward principle”, this hybrid model makes a distinction between the reuse of generic terms that are more or less established and other reused terms.

Terms that are genuinely generic form a permanent part of the NL-GEN, that is not adapted if Dutch Taxonomy partners wish to use one another's data. The NL-GEN includes the terms from the balance sheet, profit and loss report (both based on the most expansive model in the Financial Statements (Formats) Decree), cash flow overview (based on RJ 360) and the accompanying bases and general information. The objective is to keep this NL-GEN as stable as possible and to amend this as little as possible. Therefore, not every term that is used by one or more Dutch Taxonomy partners is a generic term. The Dutch Taxonomy domains have a duty of care to one another in respect of the terms included in the NL-GEN. This means that Dutch Taxonomy domains have to agree with other Dutch Taxonomy domains how amendments will be made to the generic concepts.

The other reused terms will be elaborated on further in the section "General reuse of terms in domain extensions" at the bottom of this page.


In the NL-CD taxonomy (Dutch Common Dictionary elements), among others, separate schemas are provided for terms and/or values that are issued by external organisations. This mainly relates to the ISO and UN-CEFACT code lists for currency (ISO4217), countries (ISO3166) and units (UN-Cefact66411).

External code lists and terms will possibly be supplied at different times to the Dutch Taxonomy. To give Dutch Taxonomy domains the ability to use the issued terms and values known at the specific time, the schemas in which these terms and values are incorporated are published in line with the publication schedule of the administrator of these standards. Therefore, the date of publication by the external party will be incorporated into the schema name.

The NL-CD also uses the XBRL Dimensions specification. Therefore terms will be supplied both as enumeration and members of a domain. Especially the last version can result in the parties making the request having to enter the <xbrli:unit> node in the instance by reports with the abstract domain members. By using the UN/CEFACT list (approx. 1,100 units) the correct unit is always available. For currency units, a ISO4217 schema had already been made mandatory (made available by XBRL).

The self-defined elements are based on the record databases that are available. For personal data this is supported by the Municipal Personal Records Database (GBA), for commercial data by the Basic Business Register (BBR) and for addresses by the NEN standards (Netherlands Standardisation Institute). In terms of SBR, a choice still has to be made whether the element names defined in the basic administration will be used or own SBR names. The Dutch Taxonomy uses English element names, whilst the record databases define Dutch element names. The references of these elements do refer to the definitions and documentation of the record databases.

Moreover, not all terms from the record databases will be included in the NL-CD. The only terms to be included are terms that the Dutch Taxonomy domains indicate will be used.

General reuse of terms in domain extensions

There are two ways of generally reusing terms in the Dutch Taxonomy: explicit reuse and implicit reuse.

The reuse can be considered implicit if the parties use one another's elements in presentation linkbases or in tuples. In this case the terms of other parties are directly incorporated in the own domain taxonomy. This means that this term remains in the namespace of the original party and that this namespace is imported by the party that will reuse it. This form of reuse is considered to be implicit because this relationship is not given in meta format. However, the reuse can (easily) be identified, as the prefix of this term differs from the prefix of the namespace of the relevant party creating a taxonomy.

The reuse is explicit when a Dutch Taxonomy partner that wishes to reuse a term copies an existing term from a different Dutch Taxonomy partner to its own namespace and shows this relationship in meta format using a Definition link (D-link). Definition links are discussed further in the section below.

Definition links

The XBRL specification has given a number of arc roles in this type of linkbase which show that elements are homonymous or synonymous of one another: general-special, essence-alias and similar-tuple. However, the XBRL specification does not state whether these definitions only apply to semantics (functional meaning) and/or to the technical definitions such as the datatype that is used. In the Dutch Taxonomy, the general-specific arc role has been chosen, to indicate that concepts are similar semantically, but different from one another technically.

If market parties require this, the implicit reuse of elements can be made explicit. A Dutch Taxonomy partner that wishes to reuse a term will then have to define an existing term of another Dutch Taxonomy partner in its own namespace. In that case, this Dutch Taxonomy partner also provides a Definition linkbase in which the (re)used elements, defined by another Dutch Taxonomy party, are identified in an SBR-specific arc role. This definition linkbase enables software suppliers and other interested parties to process the relationship in metaformat and to easily trace 'duplicates', without this having an adverse effects on those parties who are unable to use this information. A linkbase of this kind is NOT actively linked from a schema by means of a link:linkbaseRef. The reason for this is that when opening an entry point that discovers a linkbase of this kind, multiple required data has to be retrieved, i.e. the concepts and the relationships of the Dutch Taxonomy partner, where this would be 'loaned'.

The Dutch Taxonomy partners have NO duty of care in respect of elements that are reused by other parties outside of the GEN-base. However they DO have to actively inform (market) parties about amendments to the elements, for example, by providing versioning information with the Dutch Taxonomy.

Datatype constraints

When reusing widely shared data, it may be the case that parties agree about the semantic definition, but not about the technical impact in a datatype. Amounts and clarifications are mostly permitted 'unlimitedly' in the Dutch Taxonomy. However, if a receiving system of one party can only cope with amounts up to 1 billion, or clarifications up to 255 characters, this will have an impact on the reuse. The organisation that has the constraints cannot impose these on the other Dutch Taxonomy partners.

For example, the Tax and Customs Administration deal with a number of these constraints in their processing systems. Some of these constraints are resolved by defining additional requirements in the datatype. However, the generic element -intended for reuse- is not constrained. In this case, the Tax and Customs Administration creates its own concept, with an adapted datatype in which the permitted length is constrained. This element HAS to have the same name as the generic element, only the datatype is different. This construction will remain as is until it is possible to enforce these kinds of constraints using XBRL Formulas. To make the users of the Dutch Taxonomy aware of these situations, the Tax and Customs Administration provides a definition linkbase with which the two concepts are linked to one another by the general-special arc role (similar semantically, different technically).